What is your IQ?

It appears that he was speaking about "confidence intervals" rather than "deviation."  In any case, that's not exactly how one would use confidence intervals anyway.  The confidence interval is a set of values centered around the mean.  The radius of that interval is called the margin of error and it is dependent on the confidence level, standard deviation, and the sample size.  Suppose that this particular test has a 95% confidence interval for a sample size of 1000 with a mean of 100 and standard deviation of 15 (the last two values are characteristic of IQ tests).  Using a straightforward formula, we find that the margin of error is approximately 0.923.  This would mean that we should expect the actual mean to be between 99.077 and 100.923 (roughly 99 and 101).  I can't think of any situation in which a margin of error of five to ten percent of the mean would be considered acceptable.

I said it in my first post, "She explained me in technical words that i hardly understood", I only exptressed what I understood after an explanation in simpler terms..

Most modern psychologists don't take IQ tests seriously, or if they do it is cum grano salis.  In particular, the concept of "genius" is generally ill-defined, or in the most charitable analysis it is nebulously conceived.  Moreover, I'd be quite surprised to learn of any company that expresses any concern over the IQ of a prospective hire.

They do also, but she was talking about a determined test that is in the first page, not about all tests.

Again, this strikes me as simply bizarre.  Perhaps employment procedures differ in the southern hemisphere: this simply doesn't happen up here.

Asked to the expert in the personnel department about the test in the front page.

He explained me in technical words that i hardly understood, but to summarize that test has a deviation of + 5 to + 10% in the final result.

It appears that he was speaking about "confidence intervals" rather than "deviation."  In any case, that's not exactly how one would use confidence intervals anyway.  The confidence interval is a set of values centered around the mean.  The radius of that interval is called the margin of error and it is dependent on the confidence level, standard deviation, and the sample size.  Suppose that this particular test has a 95% confidence interval for a sample size of 1000 with a mean of 100 and standard deviation of 15 (the last two values are characteristic of IQ tests).  Using a straightforward formula, we find that the margin of error is approximately 0.923.  This would mean that we should expect the actual mean to be between 99.077 and 100.923 (roughly 99 and 101).  I can't think of any situation in which a margin of error of five to ten percent of the mean would be considered acceptable.

So if you have 150, your real IQ is probably between 135 - 142.

See above.

She knew about this short tests, are a good tool for fast exams, because a company doesn't care too much if you are a genius, they care only if your IQ is in a normal range and that can be answered by this short tests, but according to her, this one has been manipulated to boost the results, specially for people who is over 130.

Most modern psychologists don't take IQ tests seriously, or if they do it is cum grano salis.  In particular, the concept of "genius" is generally ill-defined, or in the most charitable analysis it is nebulously conceived.  Moreover, I'd be quite surprised to learn of any company that expresses any concern over the IQ of a prospective hire.

Bellow 130 the deviation doesn't exist, because  the target people that companies search for is between 110 - 130 (except for crucial positions, but those positions are normally covered by people recruited directly from the Universities or experts in a determined field, so they don't take the test), .

Again, this strikes me as simply bizarre.  Perhaps employment procedures differ in the southern hemisphere: this simply doesn't happen up here.